MAP-based estimation of the parameters of non-stationary Gaussian processes from noisy observations.
Alexander KruegerReinhold Haeb-UmbachPublished in: ICASSP (2011)
Keyphrases
- non stationary
- gaussian processes
- covariance function
- noisy observations
- hyperparameters
- gaussian process
- parameter estimation
- maximum a posteriori
- model selection
- parameter values
- gaussian process regression
- empirical mode decomposition
- gaussian process models
- maximum likelihood
- parameter settings
- incremental learning
- cross validation
- image processing
- parameter space
- regularization parameter
- image reconstruction
- markov random field